AI优化巴西物流可减15%排放，中资货运企业短期降本窗口开启

Artificial intelligence is moving beyond process automation to become an ally in transportation decarbonization by enabling more efficient operations, lower fuel use, better logistics planning, and integration among different modes. In a sector still largely dependent on fossil-fuel vehicles, experts believe the technology can begin reducing emissions even before fleets are electrified on a large scale by improving operational efficiency, reducing waste, and optimizing existing infrastructure. That assessment is consistent with a study released in 2025 by the Transportation Coalition, which brings together more than 50 industry associations, companies, and academic institutions. The report indicates that transportation accounts for approximately 11% of Brazil’s greenhouse-gas emissions and that road transportation generates more than 90% of those emissions. For road transportation, the study identifies operational efficiency as one way to reduce emissions in the short term, alongside expanding biofuel use, electrification, and changing the country’s transportation mix. This includes using artificial intelligence to improve freight routes and lower emissions. A 2025 report by the World Economic Forum and consulting firm McKinsey reached a similar conclusion. It found that AI cannot replace the structural changes required to decarbonize transportation, but has greater potential than any other technology to reduce emissions in the short term by making better use of existing assets. According to the report, titled “Smart Transportation, Greener Future: Artificial Intelligence as a Catalyst for Decarbonizing Global Logistics,” adopting these technologies could reduce global logistics emissions by as much as 15% through operational-efficiency improvements alone. Environment Minister João Paulo Capobianco tells Valor that in a country such as Brazil, where logistics costs have a major impact on production, even modest efficiency gains can generate economic benefits. They also produce environmental gains because the less energy consumed per tonne moved or passenger carried, the lower the emissions associated with economic activity. In this context, Capobianco views artificial intelligence as one of several solutions for decarbonizing the sector. “Electrification will be one of the pillars of transportation decarbonization, but its results will be much greater when combined with other technological transformations. Artificial intelligence is one of them. Its primary role is not to replace electrification but to increase its efficiency and integrate different solutions to reduce emissions,” he says. According to Capobianco, this combination is particularly relevant in Brazil because of the country’s predominantly renewable electricity mix and the progress made with low-carbon fuels, which allow different solutions to be used according to the characteristics of each transportation mode. Decarbonizing transportation is unlikely to result from a single technology, he adds, and will depend on the ability to integrate different approaches into a more efficient system. “This convergence can accelerate the transition to a low-carbon economy,” he explains. AI’s role is expected to expand as transportation, energy, logistics, and urban-planning systems become increasingly integrated. According to Capobianco, the government must create conditions for these innovations to develop at scale through regulatory predictability, financing mechanisms, research support, and an environment conducive to private investment. “The greater the integration, the greater the ability to reduce energy use, emissions, and operating costs. That is why, rather than measuring artificial intelligence’s direct impact, it is more important to understand its role as an enabling technology. It improves the efficiency of a range of solutions that, when combined, can significantly accelerate the decarbonization of the transportation sector,” Capobianco notes. Artificial intelligence is already part of the operational routines of major transportation companies, says Fernanda Rezende, executive director of the National Transportation Confederation (CNT). The technology is no longer limited to automating administrative processes and is now being applied directly to logistics operations. AI is used, for example, to optimize routes and load cubing—the process of determining the best way to arrange goods inside vehicles to maximize the use of available space. This makes it possible to carry more freight with the same number of vehicles and reduce unnecessary trips, particularly in less-than-truckload operations. “AI can be used to calculate loads and assemble them more efficiently inside a truck, increasing the sector’s efficiency,” she explains. The technology is also used to analyze vehicle data through telemetry, which monitors performance and fuel use in real time. It supports predictive maintenance by identifying malfunctions and component wear before they impair vehicle performance, while also expanding the monitoring capacity of control centers. According to Rezende, these capabilities can further improve fleet operations and help reduce fuel use. Rezende also cites advances in autonomous trucks operating in controlled environments. However, she believes deploying the technology on highways still faces obstacles related to the condition of Brazil’s infrastructure. “These levels of automation are already a reality. Technology is already being used to improve safety and increase transportation efficiency,” she says. Artificial intelligence applications are not limited to road transportation. They are also being incorporated into other modes, including aviation. In the port and waterway sector, artificial intelligence is primarily being used to improve operational efficiency, according to Frederico Dias, managing director of the National Waterway Transportation Agency (Antaq). The technology can support route planning, the monitoring of weather and oceanographic conditions, arrival forecasts, port-flow management, and reductions in idle time for vessels, equipment, and vehicles. “These efficiency gains have a direct impact on decarbonization because they reduce fuel use, energy use, and emissions,” he says. Dias says Antaq is paying particular attention to the potential of monitoring systems, data integration, and geospatial intelligence to support spatial analysis, satellite-image processing, the identification of operational patterns, and regulatory decision-making. Recent international studies indicate that increased digitalization is associated with lower costs and energy use while enabling real-time monitoring, alerts, and more intelligent management of operations, he adds. “AI can also be used to predict container dwell times and vessel delays, improve stacking, and automate equipment. The central bottleneck is data integration: without reliable, standardized, and shared data, AI cannot realize its full environmental and operational potential,” Dias says. Li Weigang, a professor in the University of Brasília’s Department of Computer Science and coordinator of TransLab, says the Brazilian Artificial Intelligence Plan (PBIA) establishes guidelines for using the technology to increase the country’s competitiveness and designates transportation as a priority area. According to Li, modernizing logistics infrastructure requires the use of artificial intelligence. “The plan recognizes infrastructure and mobility as priority areas. Its greatest strength is its focus on developing domestic solutions by optimizing logistics corridors, port infrastructure, and intelligent urban-traffic management to reduce carbon emissions,” he says. Li views the strategy as an opportunity to develop technologies adapted to Brazilian conditions instead of merely importing solutions. “In transportation, the PBIA focuses on four areas: technological sovereignty and domestic solutions; modernization of logistics infrastructure; urban mobility and sustainability; and safety and resilience,” he says. Translation: Todd Harkin